SEARCH · Engineering Papers

Results for “low memory”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,189 records · Page 66

All sources NASA NTRS DOE OSTI

NASA NTRSDate not supplied

Towards Accurate Predictions of Martensitic Transition Temperatures for Shape Memory Alloys from Ab Initio Simulations

Experimentally, NiTi undergoes a single martensitic phase transition around 341 K from the lowtemperature (T) monoclinic B19’ phase (P21/m) to the high-temperature cubic B2 phase (Pm3m). Theoretically, an orthorhombic B33 (Cmcm) has also been proposed as the T=0 ground state structure, although this phase has never been observed. Accurate predictions of martensitic transition temperatures (MTT) have remained elusive in part due to several well-known theoretical complexities of these systems including low temperature instabilities of the B2 phase. Recently, we proposed a rigorous thermodynamic integration approach based on ab initio simulations to resolve many of these difficulties [1,2]. However, an unsatisfying overprediction of the MTT relative to experiment (by ~100 K) means a fully quantitative theory is still lacking. In this work, we report several new developments to our method that bring first principles theory and experiment much closer into agreement. Our calculations indicate that phonon free energies at low temperature stabilizes B19’ over B33, rationalizing B19’ as the ground state down to T=0. We also find that accurate computations of the electronic free energy, i.e. the change in energy and the appearance of electronic configurational entropy due to finite temperature, is crucial to obtain accurate MTT. Incorporating these corrections results in an MTT prediction of 365 K for binary NiTi, which is in very close agreement with experiment. Our theoretical approach is expected to be a broadly applicable and predictive theory for MTT of SMAs.

Zhigang Wu↗

DOE OSTI2026

The impact of metastability on the high-pressure behavior of cerium

The structures adopted by solids under pressure are often assumed to reflect thermodynamic equilibrium, yet in many materials phase selection is strongly influenced by kinetic pathways and microstructural inheritance. Elemental cerium (Ce) exemplifies this challenge, with decades of conflicting reports describing different high-pressure crystal structures emerging under nominally identical conditions. Here we use neutron diffraction from large ( ~ 60 mm 3 ) sample volumes to follow the structural evolution in ultra-high-purity Ce during controlled pressure-temperature cycling between 85 and 295 K and up to 8 GPa. We find that the crystal structure formed at high pressure depends on the compression pathway: slow compression ( ~ 0.25 GPa hr −1 ) at room temperature favors an orthorhombic phase (α'), whereas slow ( ~ 0.25 GPa hr −1 ) and also moderately faster ( ~ 0.5 GPa hr −1 ) compression at low temperature stabilizes a pure monoclinic phase (α"). The low-temperature phase persists metastably over a wide temperature range but transforms irreversibly upon heating above ~ 280 K, or modest pressure cycling. Remarkably, the lower-pressure γ phase remains trapped far beyond the expected stability field, persisting to the highest pressures studied. These observations show that phase selection in Ce is governed by kinetics and microstructural memory rather than equilibrium thermodynamics or sample morphology alone, establishing path dependence as a defining feature of its high-pressure behavior.

Ridley, Christopher J. [Oak Ridge National Laborat↗

DOE OSTI2021

Revisiting Huffman Coding: Toward Extreme Performance on Modern GPU Architectures

Today's high-performance computing (HPC) applications are producing vast volumes of data, which are challenging to store and transfer efficiently during the execution, such that data compression is becoming a critical technique to mitigate the storage burden and data movement cost. Huffman coding is arguably the most efficient Entropy coding algorithm in information theory, such that it could be found as a fundamental step in many modern compression algorithms such as DEFLATE. On the other hand, today's HPC applications are more and more relying on the accelerators such as GPU on supercomputers, while Huffman encoding suffers from low throughput on GPUs, resulting in a significant bottleneck in the entire data processing. In this paper, we propose and implement an efficient Huffman encoding approach based on modern GPU architectures, which addresses two key challenges: (1) how to parallelize the entire Huffman encoding algorithm, including codebook construction, and (2) how to fully utilize the high memory-bandwidth feature of modern GPU architectures. The detailed contribution is fourfold. (1) We develop an efficient parallel codebook construction on GPUs that scales effectively with the number of input symbols. (2) We propose a novel reduction based encoding scheme that can efficiently merge the codewords on GPUs. (3) We optimize the overall GPU performance by leveraging the state-of-the-art CUDA APIs such as Cooperative Groups. (4) We evaluate our Huffman encoder thoroughly using six real-world application datasets on two advanced GPUs and compare with our implemented multi-threaded Huffman encoder. Experiments show that our solution can improve the encoding throughput by up to 5.0x and 6.8x on NVIDIA RTX 5000 and V100, respectively, over the state-of-the-art GPU Huffman encoder, and by up to 3.3x over the multi-thread encoder on two 28-core Xeon Platinum 8280 CPUs.

Tian, Jiannan↗

DOE OSTI2020

TuckerMPI: A Parallel C++/MPI Software Package for Large-scale Data Compression via the Tucker Tensor Decomposition

With this study, our goal is compression of massive-scale grid-structured data, such as the multi-terabyte output of a high-fidelity computational simulation. For such data sets, we have developed a new software package called TuckerMPI, a parallel C++/MPI software package for compressing distributed data. The approach is based on treating the data as a tensor, i.e., a multidimensional array, and computing its truncated Tucker decomposition, a higher-order analogue to the truncated singular value decomposition of a matrix. The result is a low-rank approximation of the original tensor-structured data. Compression efficiency is achieved by detecting latent global structure within the data, which we contrast to most compression methods that are focused on local structure. In this work, we describe TuckerMPI, our implementation of the truncated Tucker decomposition, including details of the data distribution and in-memory layouts, the parallel and serial implementations of the key kernels, and analysis of the storage, communication, and computational costs. We test the software on 4.5 and 6.7 terabyte data sets distributed across 100 s of nodes (1,000 s of MPI processes), achieving compression ratios between 100 and 200,000×, which equates to 99--99.999% compression (depending on the desired accuracy) in substantially less time than it would take to even read the same dataset from a parallel file system. Moreover, we show that our method also allows for reconstruction of partial or down-sampled data on a single node, without a parallel computer so long as the reconstructed portion is small enough to fit on a single machine, e.g., in the instance of reconstructing/visualizing a single down-sampled time step or computing summary statistics. The code is available at https://gitlab.com/tensors/TuckerMPI.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2016

HPC-NMF: A High-Performance Parallel Algorithm for Nonnegative Matrix Factorization

NMF is a useful tool for many applications in different domains such as topic modeling in text mining, background separation in video analysis, and community detection in social networks. Despite its popularity in the data mining community, there is a lack of efficient distributed algorithms to solve the problem for big data sets. We propose a high-performance distributed-memory parallel algorithm that computes the factorization by iteratively solving alternating non-negative least squares (NLS) subproblems for $\WW$ and $\HH$. It maintains the data and factor matrices in memory (distributed across processors), uses MPI for interprocessor communication, and, in the dense case, provably minimizes communication costs (under mild assumptions). As opposed to previous implementation, our algorithm is also flexible: It performs well for both dense and sparse matrices, and allows the user to choose any one of the multiple algorithms for solving the updates to low rank factors $\WW$ and $\HH$ within the alternating iterations.

Kannan, Ramakrishnan↗

DOE OSTI2017

Exploring Electric Polarization Mechanisms in Multiferroic Oxides

Multiferroic oxides are a class of systems which exhibit coupling between the electrical polarization and the magnetization. These materials show promise to lead to devices in which ferromagnetic memory can be written with magnetic fields or magnetic bits can be written by an electric field. The work conducted in our research focuses on single phase materials. We studied the detailed coupling of the spin and lattice correlations in these systems. In the first phase of the proposal, we explored the complex spin spiral systems and low temperature behavior of hexagonal layered REMnO 3 (RE= rare earth, Y and Sc) system following the detailed structural changes which occurred on crossing into the magnetic states. The techniques were applied to other layered materials such as superconductors and thermoelectric where the same layered motif exists. The second phase of the proposal focused on understanding the mechanisms involved in the onset high temperature ferroelectricity ion hexagonal REMnO 3 and at low temperature in E-Type magnetic ordered perovskite REMnO 3 . We wsynthesized preovskite small A site multiferroics by high pressure and high temperature methods. Detailed measurement of the structural properties and dynamics were conducted over a range of length scales from atomic to mesoscopic scale using, x-ray absorption spectroscopy, x-ray diffuse scattering, x-ray and neutron pair distribution analysis and high resolution x-ray diffraction. Changes in vibration modes which occur with the onset of polarization were probed with temperature and pressure dependent infrared absorption spectroscopy. In addition the orthorhombic system (small radius RE ions) which is believed to exhibit electronically driven ferroelectricity and is also not understood was examined. The multiple length scale synchrotron based measurements may assist in developing more detailed models of these materials and possibly lead to device applications. The experimental work was complemented by density functional methods to determine the magnetic ground states and ab initio molecular dynamics methods (AIMD) to determine the high temperature structures. Simulation were carried out on supercomputers at the National Energy Research Scientific Computing Center (NERSC). An important contribution of this work was the training of graduate students and postdoctoral researchers in materials synthesis, high pressure methods and synchrotron based spectroscopy and x-ray scattering techniques.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2020

CRADA Final Report: Open Microgrid Platform

Current microgrid control technology is mostly proprietary, expensive, and inflexible. An open-source, vendor-neutral, publish-subscribe controller will significantly lower the barrier to entry for new technologies and market entrants: inverter, battery, and load control manufacturers; cloud and services providers; and energy aggregators. An OpenFMB controller can significantly reduce the acquisition, integration, and ongoing operating and maintenance costs and increase revenue opportunities for energy producers and consumers. It allows distributed energy resource (DER) owners to replace individual DER components as richer-function, lower-cost devices become available or to incorporate new forecast and optimization algorithms as they are developed. A standard, secure, full-function DER field controller will also provide a low-cost means for grid operators to effectively manage the variability and uncertainty of solar photovoltaics (PV). Under the terms of this CRADA, ORNL has worked with Open Energy Solutions (OES) to advance the transition of ORNL’s existing open-source microgrid controller into a platform geared toward mainstream implementation. Using a more prevalent and memory-safe programming language, Rust, OES has converted ORNL’s current generation on-grid optimization into an application more suitable for industry use. The results of the ported optimization were analyzed by inputting the same arguments in both the current-generation and next-generation optimizers and verifying that the results were equivalent.

24 POWER TRANSMISSION AND DISTRIBUTION↗

NASA NTRS1994

Improved load distribution in parallel sparse Cholesky factorization

Compared to the customary column-oriented approaches, block-oriented, distributed-memory sparse Cholesky factorization benefits from an asymptotic reduction in interprocessor communication volume and an asymptotic increase in the amount of concurrency that is exposed in the problem. Unfortunately, block-oriented approaches (specifically, the block fan-out method) have suffered from poor balance of the computational load. As a result, achieved performance can be quite low. This paper investigates the reasons for this load imbalance and proposes simple block mapping heuristics that dramatically improve it. The result is a roughly 20% increase in realized parallel factorization performance, as demonstrated by performance results from an Intel Paragon system. We have achieved performance of nearly 3.2 billion floating point operations per second with this technique on a 196-node Paragon system.

Rothberg, Edward↗

NASA NTRS2000

First-Generation Jet Propulsion Laboratory "Hockey-Puck" Free-Flying Magnetometers for Distributed In-Situ Multiprobe Measurement of Current Density Filamentation in the Northern Auroral Zone: Enstrophy Mission

The sub-orbital rocket mission was a collaborative project between the University of New Hampshire, Cornell University, and the Jet Propulsion Laboratory (JPL) to study filamentation phenomena in the northern Auroral zone. The Enstrophy mission test flies the JPL Free-Flying Magnetometer (FFM) concept. The FFM technology development task has been funded by NASA develop miniaturized, low-power, integrated "sensorcrafts". JPL's role was to design, integrate, test, and deliver four FFMs for deployment from the sounding rocket, allowing a unique determination of curl-B. This provides a direct measurement of magnetic-field-aligned current density along the rocket trajectory. A miniaturized three-axis fluxgate magnetometer was integrated with a 4-channel 22-bit sigma-delta Analog to Digital Converter (ADC), four temperature sensors, digital control electronics, seven (Li-SOCl2) batteries, two (4 deg x 170 deg field of view) sun-sensors, a fan-shaped-beam laser diode beacon, a (16 MHz) stable Temperature Compensated Crystal Oscillator (TCXO) clock, Radio Frequency (RF) communication subsystem, and an antenna for approximately 15 minutes of operation where data was collected continuously and transmitted in three (3) bursts (approximately 26 seconds each) to ground station antennas at Poker Flat, Alaska. FFMs were stowed within two trays onboard the rocket during the rocket launch and were released simultaneously using the spinning action of the rocket at approximately 300 km altitude (approximately 100 sec. into the flight). FFMs were deployed with spin rate of approximately 17 Hz and approximately 3 m/sec linear velocity with respect to the rocket. For testing purposes while the rocket was in the launch pad and during flight prior to release of FFMs from the rocket, commands (such as "power on", "test", "flight", "power off', and clock "Reset" signal) were transmitted via a infrared Light Emitting Diode to an infrared detector in the FFM. Special attention was paid to low magnetic signature electronic design and choice of materials in packaging. The miniaturized fluxgate magnetometers had a range of 1-60000 nT with 0.1% full-scale linearity. The frequency range of interest for magnetic measurement was 10 mHz - 50 Hz. Digital data from the magnetometer's three axes were placed in a 4MB Static Random Access Memory (SRAM) in data packages (frames) formatted together with time tags and frame ID. After a specified time was elapsed, the data were Viterbi encoded and transmitted at a rate of 100 kbps (BPSK). Each of the four FFMs transmitted at different frequency. These carrier frequencies were in the range of 2200-2300 MHz. The antenna was a single patch on a high dielectric constant substrate covering one end-plate of the hockey-puck-sized unit. The local clocks aboard the FFMs were reset at the start of the mission and stayed synchronized within 3 msec during the mission. Position of each FFM with respect to the rocket is calculated by the knowledge of its release velocity (measured at exit point of the FFM launcher tract) providing an accuracy of 1 m over the maximum range of 3 km. Spatial and temporal nature of observants can be separated to within 3 m in space or 3 msec time interval.

Javadi, H.↗

DOE OSTI2021

Revisiting Huffman Coding: Toward Extreme Performance on Modern GPU Architectures

Today's high-performance computing (HPC) applications are producing vast volumes of data, which are challenging to store and transfer efficiently during the execution, such that data compression is becoming a critical technique to mitigate the storage burden and data movement cost. Huffman coding is arguably the most efficient Entropy coding algorithm in information theory, such that it could be found as a fundamental step in many modern compression algorithms such as DEFLATE. On the other hand, today's HPC applications are more and more relying on the accelerators such as GPU on supercomputers, while Huffman encoding suffers from low throughput on GPUs, resulting in a significant bottleneck in the entire data processing. In this paper, we propose and implement an efficient Huffman encoding approach based on modern GPU architectures, which addresses two key challenges: (1) how to parallelize the entire Huffman encoding algorithm, including codebook construction, and (2) how to fully utilize the high memory-bandwidth feature of modern GPU architectures. The detailed contribution is fourfold. (1) We develop an efficient parallel codebook construction on GPUs that scales effectively with the number of input symbols. (2) We propose a novel reduction based encoding scheme that can efficiently merge the codewords on GPUs. (3) We optimize the overall GPU performance by leveraging the state-of-the-art CUDA APIs such as Cooperative Groups. (4) We evaluate our Huffman encoder thoroughly using six real-world application datasets on two advanced GPUs and compare with our implemented multithreaded Huffman encoder. Experiments show that our solution can improve the encoding throughput by up to 5.0× and 6.8× on NVIDIA RTX 5000 and V100, respectively, over the state-of-the-art GPU Huffman encoder, and by up to 3.3× over the multithread encoder on two 28-core Xeon Platinum 8280 CPUs.

Tian, Jiannan↗

DOE OSTI2022

Probabilistic Nanomagnetic Memories for Uncertain and Robust Machine Learning

This project evaluated the use of emerging spintronic memory devices for robust and efficient variational inference schemes. Variational inference (VI) schemes, which constrain the distribution for each weight to be a Gaussian distribution with a mean and standard deviation, are a tractable method for calculating posterior distributions of weights in a Bayesian neural network such that this neural network can also be trained using the powerful backpropagation algorithm. Our project focuses on domain-wall magnetic tunnel junctions (DW-MTJs), a powerful multi-functional spintronic synapse design that can achieve low power switching while also opening the pathway towards repeatable, analog operation using fabricated notches. Our initial efforts to employ DW-MTJs as an all-in-one stochastic synapse with both a mean and standard deviation didn’t end up meeting the quality metrics for hardware-friendly VI. In the future, new device stacks and methods for expressive anisotropy modification may make this idea still possible. However, as a fall back that immediately satisfies our requirements, we invented and detailed how the combination of a DW-MTJ synapse encoding the mean and a probabilistic Bayes-MTJ device, programmed via a ferroelectric or ionically modifiable layer, can robustly and expressively implement VI. This design includes a physics-informed small circuit model, that was scaled up to perform and demonstrate rigorous uncertainty quantification applications, up to and including small convolutional networks on a grayscale image classification task, and larger (Residual) networks implementing multi-channel image classification. Lastly, as these results and ideas all depend upon the idea of an inference application where weights (spintronic memory states) remain non-volatile, the retention of these synapses for the notched case was further interrogated. These investigations revealed and emphasized the importance of both notch geometry and anisotropy modification in order to further enhance the endurance of written spintronic states. In the near future, these results will be mapped to effective predictions for room temperature and elevated operation DW-MTJ memory retention, and experimentally verified when devices become available.

77 NANOSCIENCE AND NANOTECHNOLOGY↗

DOE OSTI2014

Joint contributions of Ag ions and oxygen vacancies to conducting filament evolution of Ag/TaO{sub x}/Pt memory device

The electroforming and resistive switching behaviors in the Ag/TaO{sub x}/Pt trilayer structure are investigated under a continual change of temperatures between 300 K and 100 K to distinguish the contributions of Ag ions and oxygen vacancies in developing of conducting filaments. For either electroforming or resistive switching, a significantly higher forming/set voltages is needed as the device is operated at 100 K, as compared to that observed when operating at 300 K. The disparity in forming/set voltages of Ag/TaO{sub x}/Pt operating at 300 K and 100 K is attributed to the contribution of oxygen vacancies, in addition to Ag atoms, in formation of conducting filament at 100 K since the mobilities of oxygen vacancies and Ag ions become comparable at low temperature. The presence of oxygen vacancy segment in the conducting filament also modifies the reset current from a gradually descending behavior (at 300 K) to a sharp drop (at 100 K). Furthermore, the characteristic set voltage and reset current are irreversible as the operation temperature is brought from 100 K back to 300 K, indicating the critical role of filament constituents on the switching behaviors of Ag/oxide/Pt system.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

DOE OSTI2016

Low Dose Ionizing Radiation Modulates Immune Function

In order to examine the effects of low dose ionizing radiation on the immune system we chose to examine an amplified adaptive cellular immunity response. This response is Type IV delayed-type hypersensitivity also called contact hypersensitivity. The agent fluorescein isothiocyanate (FITC) is a low molecular weight, lipophilic, reactive, fluorescent molecule that can be applied to the skin where it (hapten) reacts with proteins (carriers) to become a complete antigen. Exposure to FITC leads to sensitization which is easily measured as a hypersensitivity inflammatory reaction following a subsequent exposure to the ear. Ear swelling, eosinophil infiltration, immunoglobulin E production and cytokine secretion patterns characteristic of a “Th2 polarized” immune response are the components of the reaction. The reaction requires successful implementation of antigen processing and presentation by antigen presenting Langerhans cells, communication with naïve T lymphocytes in draining lymph nodes, expansion of activated T cell clones, migration of activated T cells to the circulation, and recruitment of memory T cells, macrophages and eosinophils to the site of the secondary challenge. Using this model our approach was to quantify system function rather than relying only on indirect biomarkers of cell. We measured the FITC-induced hypersensitivity reaction over a range of doses from 2 cGy to 2 Gy. Irradiations were performed during key events or prior to key events to deplete critical cell populations. In addition to quantifying the final inflammatory response, we assessed cell populations in peripheral blood and spleen, cytokine signatures, IgE levels and expression of genes associated with key processes in sensitization and elicitation/recall. We hypothesized that ionizing radiation would produce a biphasic effect on immune system function resulting in an enhancement at low doses and a depression at higher doses and suggested that this transition would occur in the dose range of 5 to 50 cGy.

63 RADIATION, THERMAL, AND OTHER ENVIRON. POLLUTAN↗

DOE OSTI2015

An interface for the direct coupling of small liquid samples to AMS

We describe the moving wire interface attached to the 1-MV AMS system at LLNL’s Center for Accelerator Mass Spectrometry for the analysis of nonvolatile liquid samples as either discrete drops or from the direct output of biochemical separatory instrumentation, such as high-performance liquid chromatography (HPLC). Discrete samples containing at least a few 10 s of nanograms of carbon and as little as 50 zmol 14 C can be measured with a 3–5% precision in a few minutes. The dynamic range of our system spans approximately 3 orders in magnitude. Sample to sample memory is minimized by the use of fresh targets for each discrete sample or by minimizing the amount of carbon present in a peak generated by an HPLC containing a significant amount of 14 C. As a result, liquid sample AMS provides a new technology to expand our biomedical AMS program by enabling the capability to measure low-level biochemicals in extremely small samples that would otherwise be inaccessible.

14C↗

NASA NTRS2006

Micro-Inspector Avionics Module (MAM): A Self-Contained Low Power, Reconfigurable Avionics Platform for Small Spacecrafts and Instruments

This paper describes development of a radiation tolerant, low power, reconfigurable avionics module aimed at meeting the avionics needs of the JPL Micro-Inspector spacecraft. This module represents a complete avionics system, consisting of two PowerPC 405 CPUs embedded within a reconfigurable FPGA fabric of over 8 Million logic gates, 64MB of EDAC protected Flash storage and 128MB of EDAC protected DDR SDRAM or SDRAM memories, along with FPGA SEU mitigation logic, and all necessary power conversion. Processor SEU mitigation is achieved by running the two processors in a lock-step and compare configuration. All of these building blocks are integrated into a double sided circuit board that takes as little as 6 square inches of board space. This module can be embedded into a user system as part of a bigger circuit assembly or as a self contained module. This module is being developed as part of a JPL led Micro-Inspector Program, funded by NASA ESMD aimed at producing a 10Kg micro spacecraft.

configurable↗

NASA NTRS1973

Low cost automated precise time measurement system

The Aerospace Guidance and Metrology Center (AGMC) has the responsibility for the dissemination of Precise Time and Time Interval (PTTI) to Air Force timing systems requiring microsecond time. In order to maintain traceability to the USNO Master Clock in Washington D.C., and accomplish efficient logging of time and frequency data on individual precision clocks, a simple automatic means of acquiring precise time has been devised. The Automatic Time Interval Measurement System (ATIMS) consists of a minicomputer (8K Memory), teletype terminal, electronic counter, Loran C receiver, time base generator and locally-manufactured relay matrix panel. During the measurement process, the computer controls the relay matrix which selects for comparison 13 atomic clocks against a reference clock and the reference versus Loran C. Through use of the system teletype, the operator is able to set the system clock (hours, minutes and seconds), examine and/or modify all clock data and constants, and set measurement intervals. This is done in a conversational manner. A logic flow diagram, system schematic, source listing and software components are included in the presentation.

Alpert, A.↗

NASA NTRS1978

A video display interface for the LORAN-C navigation receiver development system

A microprocessor-based development system was designed and fabricated for prototype test of navigation receiver designs. During use of this system in the development of low-cost LORAN-C receiver/processor concepts, the limitations of the integral KIM-1 display were severely felt. It was to augment this numerical display that the video character display was produced. The circuit design presented meets the need for a flexible-format display capable of driving a small standard video monitor with only minimal demands upon microprocessor memory and MPU cycles.

Fischer, J. P.↗

NASA NTRS2003

Micro-Ball-Lens Optical Switch Driven by SMA Actuator

The figure is a simplified cross section of a microscopic optical switch that was partially developed at the time of reporting the information for this article. In a fully developed version, light would be coupled from an input optical fiber to one of two side-by-side output optical fibers. The optical connection between the input and the selected output fiber would be made via a microscopic ball lens. Switching of the optical connection from one output fiber to another would be effected by using a pair of thin-film shape-memory-alloy (SMA) actuators to toggle the lens between two resting switch positions. There are many optical switches some made of macroscopic parts by conventional fabrication techniques and some that are microfabricated and, hence, belong to the class of microelectromechanical systems (MEMS). Conventionally fabricated optical switches tend to be expensive. MEMS switches can be mass-produced at relatively low cost, but their attractiveness has been diminished by the fact that, heretofore, MEMS switches have usually been found to exhibit high insertion losses. The present switch is intended to serve as a prototype of low-loss MEMS switches. In addition, this is the first reported SMA-based optical switch. The optical fibers would be held in V grooves in a silicon frame. The lens would have a diameter of 1 m; it would be held by, and positioned between, the SMA actuators, which would be made of thin films of TiNi alloy. Although the SMA actuators are depicted here as having simple shapes for the sake of clarity of illustration, the real actuators would have complex, partly net-like shapes. With the exception of the lens and the optical fibers, the SMA actuators and other components of the switch would be made by microfabrication techniques. The components would be assembled into a sandwich structure to complete the fabrication of the switch. To effect switching, an electric current would be passed through one of the SMA actuators to heat it above its transition temperature, thereby causing it to deform to a different "remembered" shape. The two SMA actuators would be stiff enough that once switching had taken place and the electrical current was turned off, the lens would remain latched in the most recently selected position. In a test, the partially developed switch exhibited an insertion loss of only -1.9 dB and a switching contrast of 70 dB. One the basis of prior research on SMA actuators and assuming a lens displacement of 125 m between extreme positions, it has been estimated that the fully developed switch would be capable of operating at a frequency as high as 10 Hz.

Yang, Eui-Hyeok↗